Letter in response to Black et al. (2020): Authors’ Reply

To the editor; We are grateful to Trindade et al. 1 for their interest in our paper, which examined the factor structure of the visceral sensitivity index (VSI) among 811 individuals with Rome IV irritable bowel syndrome (IBS) from the UK.2 The VSI is a measure of gastrointestinal symptomspecific anxiety that had not been validated in the UK population. The original validation study from the USA reported an unifactorial structure,3 whereas our analysis identified three separate factors. Moreover, we found no evidence that VSI mediated the relationship between measures of generalized anxiety and symptom severity in IBS. In their analysis of 424 Swedish patients with IBS, Trindade et al. report a singlefactor structure in agreement with the original VSI study.1 They discuss that cultural and linguistic differences between studies may have influenced the responses to the VSI, but this seems unlikely. The original VSI was developed in English, and while differences in spelling and usage exist between the UK and the USA, there is nothing that should affect understanding of the VSI in the UK population. Whether Trindade et al. translated the VSI into Swedish is unclear. However, this is unlikely to explain the differences between their results and ours, given the similarity between their findings and those from the USA. The cultural concepts discussed in the VSI, such as eating out in restaurants, will be familiar to people from any Westernized society. Consequently, methodological differences between the studies are likely to be of greater relevance to understanding the contrasting results. The original validation study involved only 96 individuals who selfidentified with IBS, of whom 85% met Rome II criteria.3 We also recruited individuals who selfidentified with IBS, but our study cohort was much larger, and we only included those who met diagnostic criteria for IBS. Moreover, we defined IBS using the current Rome IV criteria.4 These are more specific for diagnosing IBS compared with previous iterations,5,6 but they also select for patients who have more severe symptoms and greater psychological comorbidity.7,8 This may explain why analysis of the VSI in our study yielded a different factor structure than was originally described. Trindade et al. recruited patients referred to a tertiary center among whom reporting gastrointestinal symptomspecific anxiety may conceivably be more homogeneous. Furthermore, it is unclear how they defined IBS in their study. Importantly, their results are not fully published, limiting comparison with our study; however, we agree that further work to validate and refine the VSI is warranted.